Influenza Virologic Surveillance: What Is the “Right Size” for New Jersey?

BACKGROUND:   The New Jersey Department of Health (NJDOH) participated in a pilot project with the Association of Public Health Laboratories to review and evaluate the Influenza Virologic Surveillance Right Size Roadmap document.  The purpose of the Roadmap is to provide guidelines to evaluate existing influenza virologic surveillance, develop best practices and identify areas for improvement.  Right Size goals include situational awareness (identifying the start of influenza season, circulating strains, and spread of influenza) and detecting rare/novel events. NJDOH evaluated its influenza surveillance systems to determine whether Right Size goals were achievable.

METHODS:   NJDOH evaluated surveillance systems using virologic and syndromic data from five seasons (2008-2009 through 2012-2013). Collected data were compared to the Roadmap and Sample Size (SS) Calculators, using inputs for optimal and minimum surveillance as defined by this document. 

RESULTS:   Using the SS Calculator, it is estimated that NJ should test 136 (minimum 74) unscreened specimens weekly for the optimal level of situational awareness using molecular detection (e.g., rRT-PCR). To detect rare/novel events at a national level, 17 (12) prescreened specimens should be tested in low flu season, and 59 (42) in high season. Data review revealed that NJ Public Health and Environmental Laboratories (PHEL) tests both unscreened and prescreened positive specimens using rapid influenza testing (i.e., immunoassays).  NJ approached the minimum for situational awareness only at peaks of severe seasons (February 2011, January 2013). PHEL tested specimens at minimum or higher levels for rare/novel event surveillance in 67 of 165 (40.6%) influenza season weeks. NJ consistently met estimates during influenza season peaks. Other Right Size criteria (timeliness, representativeness, and acceptability) were evaluated and found to meet the goals.  For timeliness, the mean time between specimen collection and reporting was 5.3 days. For representativeness, all regions were overall well represented.   Although laboratories are required to report culture and rRT-PCR positive samples, a survey indicated underreporting.

CONCLUSIONS:   NJ has a robust system of influenza surveillance, drawing information from multiple sources. According to the Roadmap, PHEL does not test the recommended minimum number of specimens for situational awareness, but does meet the recommended minimum for rare/novel event detection during season peaks.  Additionally, PHEL tests a combination of prescreened and unscreened specimens, making direct comparison with SS Calculator estimates problematic. Further investigation is recommended to determine how to utilize existing data to monitor influenza and approach Right Size sampling needs. NJDOH should improve communication with non-PHEL laboratories to ensure reporting.